Introduction to Quantitative Political Research

PSCI 3300.003 Political Science Research Methods

A. Jordan Nafa

University of North Texas

August 30th, 2022

Introduction

  • Quantitative Social Science is the use of quantitative data to study, analyze, or predict social and political phenomena

  • Outside of academic settings, this is more commonly known as data science

    • The logic of quantitative social inquiry we will cover in this class is applied in academic social sciences, government, non-profits, and in the private sector
  • This course introduces students to the logic of causal inference and the tools quantitative social scientists use to study social, political, and economic phenomena

Course Staff

  • Instructor: A. Jordan Nafa

    • Bachelor of Science in Government, Texas Woman’s University (2014-2018)

    • PhD Candidate, UNT Department of Political Science (2018-Present)

    • Expertise in applied Bayesian statistics, causal inference, and women in politics

  • Teaching Assistant: Sung Min Yun

    • PhD Student, UNT Department of Political Science (2021-Present)

Office Hours and Communication

  • Instructor: A. Jordan Nafa (adamnafa@my.unt.edu)

    • Office: Wooten Hall 171

    • Office Hours: Tuesday and Thursday 12:30-2:00 PM or by Appointment

  • Teaching Assistant: Sung Min Yun (sungminyun@my.unt.edu)

    • Office: Wooten Hall 174A

    • Office Hours: Wednesday 2:00-5:00 PM

  • We will try to respond to emails within 48 hours Monday-Friday

    • See the syllabus for details on the policy about not responding to emails asking questions that are answered by the syllabus

Why Take this Course?

  • PSCI 3300 is a required course for political science majors so you’ll have to take it at some point

  • The skills taught in this course are standard across an increasing number of industries and fields of research

    • A basic knowledge of statistics, programming, and quantitative reasoning prepares you for a range of future career opportunities or graduate school
  • Contemporary political science is a primarily quantitative field, so understanding the tools of the trade will help you understand and critique the things you read in your other classes

  • No exams!

Teaching Philosophy

  • Math for the sake of math is an unfortunately common but largely unproductive way to teach quantitative social science

  • A college-level course in fundamentals of computer programming or elementary statistics may be helpful but is not required

  • I assume you can add, subtract, multiply, divide, and have a basic understanding of order of operations–I am sure some of you will prove me wrong

  • Our focus in this course will be primarily on how to apply, interpret, and evaluate analyses in political science rather than on the statistical theory behind them

  • This course is taught from a primarily Bayesian perspective but we’ll cover some recent developments in frequentist estimation approaches as well

Teaching Philosophy

  • Though there is some introductory statistics required, in practice that means less of this \[\text{Normal}(y|\mu,\sigma) = \frac{1}{\sqrt{2 \pi}\, \sigma} \exp\left( - \, \frac{1}{2}\left(\frac{y -\mu}{\sigma} \right)^2 \right)\]

  • And more hands-on stuff like this

## Simulate 10,000 random draws from a standard normal dist
std_norm <- rnorm(n = 10e3, mean = 0, sd = 1)

## Print a summary
summary(std_norm)
     Min.   1st Qu.    Median      Mean   3rd Qu.      Max. 
-4.130991 -0.679516 -0.001870  0.006282  0.691634  4.395782 

Structure of the Course

  • Part I: Fundamentals of Political Research

    • Research questions, surveying literature, basic logic of cause and effect, theory building, and measurement
  • Part II: Foundations of Quantitative Social Science

    • Foundations of applied statistics, probability theory, Bayesian inference, and applied regression
  • Part III: Research Design and Causal Inference

    • Experimental design, hypothesis testing and theory evaluation, strategies for causal inference with observational data, and the limits of magical thinking

Required Course Materials

  • There is one required textbook for this course which can be purchased from the UNT campus bookstore or Amazon

    • Bueno de Mesquita, Ethan, and Anthony Fowler. Thinking Clearly with Data: A Guide to Quantitative Reasoning and Analysis. Princeton University Press, 2021. ISBN: 9780691214351

  • Although it is not required, I also recommend obtaining a copy of

    • Andrew Gelman, Jennifer Hill, and Aki Vehtari. Regression and Other Stories. Cambridge University Press, 2021. ISBN: 9781107676510

  • All additional readings for the course are provided via Canvas

Technical Requirements

  • To complete the requirements for this course you will need access to a laptop or desktop computer with a stable internet connection

  • I assume you have a working knowledge of computers including

    • How to download, save, move, and locate files

    • How to install software on your computer

  • If you do not have access to a personal computer that meets the minimum requirements necessary to run R, you can access both R and RStudio from computers on the UNT campus

    • The computers in the political science lab in Wooten Hall 173 should have the most recent version of R, RStudio, and RTools
  • If you expect to have difficulty meeting the technical requirements for this course you need to let me know immediately

Class Meetings

  • Tuesday/Thursday from 9:30-10:50 AM

  • Part of our course time will be spent on lecture and answering questions and the other half will be spent working with data in R

  • You should come to class having done the readings on the syllabus for that day so we can discuss anything that isn’t clear or appears confusing

    • This is not a course you can skate through without ever reading anything and if you try to do this you will very likely fail.

    • Advice on managing reading loads in this course and others during class on Thursday

  • It should go without saying, but do not come to class if you are sick

Problem Sets

  • You will complete five problem sets which will account for 40% of your overall grade and ask you to apply the things we cover in class using real or simulated data
Assignment Topic Due Date
0 Getting Started with R and Markdown September 5th
1 Data Wrangling and Visualization September 25th
2 Probability and Descriptive Statistics October 9th
3 Regression and Inferential Statistics October 30th
4 Observational Causal Inference November 20th

Final Project

  • Develop a research question relevant to political science or public policy, identify the necessary source(s) of data, and undertake an analysis using R to answer the question

  • You have the option to work in groups of up to 4 people but you must provide written notice with the names of each student in the group no later than Friday, September 23rd

  • Broken down into a series of smaller assignments over the course of the semester

    • Examples of what each of these are expected to look like and document templates will be provided via Canvas
  • Accounts for 50% of your overall grade; 70% from 8-12 page research paper and 30% from a final presentation

Course Participation

  • Remaining 10% of your grade is course participation

  • Showing up to class, asking relevant questions, and completing occasional in-class activities

  • Gives me wiggle room to pad grades at the end of the semester

  • Might also have in-class extra credit opportunities on days when attendance is scarce, but no promises

Additional Details

  • Citations are expected to conform to the style of the American Political Science Association

  • Zero tolerance policy for plagiarism of any kind

    • Penalties range from a 0 for a specific assignment to an automatic failing grade in the course depending on severity
  • We will cover how to automatically generate citations and manuscripts in the correct format and keep track of sources in class next Tuesday

Getting Started with R

  • Detailed instructions for installing R, RTools/Xcode, RStudio, and configuring Stan are provided on the course’s Canvas page

  • We will also make use of the probabilistic programming language Stan via it’s R interfaces brms and rstanarm during parts of this course

  • If you have difficulty getting these installed on your personal computer please come by one of our office hours or make an appointment for assistance

  • First problem set due Monday, September 5th to verify you have successfully completed these installations

What Should You do This Week?

  • Read the full syllabus for more details about the course and important deadlines

  • Follow the instructions on Canvas for installing R, RStudio, and Stan

    • We will go over any issues you have getting these installed on Thursday
  • Start thinking about an empirical political research question (More on this next class)